The human brain represents one of the most metabolically active organs with a highly efficient ability to extract glucose as the primary currency for energy and carbon source. In particular, neurons are distinguished in their ability to preferentially absorb glucose from a nutrient-restricted environment through the expression of high affinity glucose transporters. The most prevalent primary brain tumor, glioblastoma, ranks among the most lethal of human cancers. Like the normal brain, glioblastomas contain cellular hierarchies with self-renewing, multi-lineage cells at the apex. These brain tumor initiating cells display therapeutic resistance, promote tumor angiogenesis, and invade into normal tissues providing rationale to model their regulation and develop targeting strategies. We recently demonstrated that brain tumor initiating cells display a marked ability to survive the reduced nutrient levels found in the neoplastic brain through the cooption of the neuronal glucose transporter, GLUT3. In contrast, non-stem cell-like tumor cells underwent cell death with nutrient restriction with a cellular plasticity towards a stem cell-like state in surviving cells. Collectivly, these studies identify a novel molecular mechanism associated with the tumor cellular hierarchy that could provide a node of fragility as targeting GLUT3 expression reduced brain tumor initiating cell self-renewal and tumor growth. Like all cancers, glioblastomas display the Warburg effect, a preferential utilization of aerobic glycolysis for energy supplies. This aerobic glycolyss frees the cells from oxygen requirements and provides a steady supply of anabolic material yet is highly glucose inefficient and requires a steady supply of glucose, suggesting a potential therapeutic target. Based on this background, we hypothesize that preferential glucose uptake shields brain tumor initiating cells from extracellular energy stress and provides an ability to these cells to occupy a diverse set of niches with different metabolic limitations. The anti-angiogenic bevacizumab has shown promise in the initial response of tumors to therapy but has failed to extend survival. Studies have suggested that angiogenic inhibitor resistance is associated with impaired vascular function and metabolic shifts that may enrich for tumor initiating cells. To investigate these potential links between cellular metabolism and the tumor hierarchy, we will dissect the interplay between brain tumor initiating cells and the tumor microenvironment. In the first aim, we will determine the role of the stem cell metabolic responses in stress resistance. In the second aim, we will interrogate the role of glucose uptake in different tumor microenvironments enriched in tumor initiating cells through the use of regional biopsies from human patients and regionally specific Glut3 modification in animal studies. Finally, we will investigate the potential synthetic lethality of targeting GLUT3 with bevacizumab or ketogenic diet therapy. We will employ a series of models derived from human glioblastomas and epilepsy tissues to lay the foundation for advanced modeling of this lethal brain disease.